[openib-general] [PATCHv2] IPoIB CM Experimental support

[Michael S. Tsirkin]

>  > > Now I'm confused. Bottom line, should I try fixing up the enabling bit ASAP,
>  > > or you don't want it in 2.6.20 anyway?
>  > 
>  > Roland, could you clarify your opinion pls?
> 
> Sorry, I thought about this a fair amount.  I think I finally ended up
> feeling that the code is just too new.  I don't think anyone other
> than you has had a chance to really look at it (I certainly haven't)
> so I think we're better off not merging it.
> 
> I know that you said -mm has limited value but I actually think just
> the build coverage is worth it.  And it is surprising how many people
> are auditing the new code that shows up in -mm so I think it will help
> a fair bit.

OK, thanks for the suggestion. I'll send something to Andrew after -rc1's out.
The memory registration speedup patches can be 2.6.20 material though, can't they?


-- 
MST